December 2023 - openstack-discuss

Re: [URGENT] - Help needed with open stack services hung
by Karl Kloppenborg 12 Dec '23

12 Dec '23

Hey team, Thanks for those who replied off list. Thankfully we figured it out, it was due to underlying resolver config mysterious mulching AAAA records instead of requesting A records… Probably goes to say, Openstack needs to have some seriously heavier logging 😃 haha. Thanks! Karl. From: Dmitriy Rabotyagov <noonedeadpunk(a)gmail.com> Date: Tuesday, 12 December 2023 at 5:35 pm To: Karl Kloppenborg <kkloppenborg(a)resetdata.com.au> Subject: Re: [URGENT] - Help needed with open stack services hung Hey, > stopping services from processing requests I assume this means that services are started and do not report any errors in logs? What happens with requests themselves? Are they answered with error or just hang? Also how environment was deployed? Some deployment tooling or manually with packages, k8s? On Tue, Dec 12, 2023, 05:30 Karl Kloppenborg <kkloppenborg(a)resetdata.com.au<mailto:kkloppenborg@resetdata.com.au>> wrote: Hi Openstack Community, Unfortunately we’ve had a weird situation arise where one of our regions has just stopped dead in it’s tracks, all management systems have been rebooted but there seems to be something stopping services from processing requests. Database, RabbitMQ Memcached have all been individually tested and I cannot find fault with them. Could any openstack consultants who are available for immediate emergency work please contact me off list. Thanks, Karl. Kind Regards, Karl Kloppenborg BCompSc, CNCF-[KCNA, CKA, CKAD], LFCE, CompTIA Linux+ XK0-004 Chief Technology Officer m: +61 437 239 565<mailto:+61%20437%20239%20565> | e: kkloppenborg(a)resetdata.com.au<mailto:kkloppenborg@resetdata.com.au> | w: https://resetdata.com [signature_2024530915][Image] This email transmission is intended only for the addressee / person responsible for delivery of the message to such person and may contain confidential or privileged information. Confidentiality and legal privilege are not waived or lost by reason of mistaken delivery to you, nor may you use, review, disclose, disseminate or copy any information contained in or attached to it. Whilst this email has been checked for viruses, the sender does not warrant that any attachments are free from viruses or other defects. You assume all liability for any loss, damage or other consequences which may arise from opening or using the attachments. If you received this e-mail in error please delete it and any attachments and kindly notify us by immediately sending an email to contact(a)resetdata.com.au<mailto:contact@resetdata.com.au>

1 0

[tripleo][ironic][wallaby] Partition Table: loop
by Alexey Kashavkin 12 Dec '23

12 Dec '23

I have an issue with nodes provisioning. All my nodes are getting a partition table - loop. I forcefully specified 'disk_label:gpt/msdos' in the capabilities properties, but it didn't help. I can't figure out at what moment it creates a partition for the entire disk size: Model: Virtio Block Device (virtblk) Disk /dev/vda: 53.7GB Sector size (logical/physical): 512B/512B Partition Table: loop Disk Flags: Number Start End Size File system Flags 1 0.00B 53.7GB 53.7GB xfs This causes the following error: Failed to check the number of primary partitions present on /dev/vda for node 8d5ed008-efea-4df7-a190-aab8274a3bd5. Error: The device /dev/vda does not have a valid MBR partition table I tried to find in the ironic documentation where exactly a partition can be partitioned to the whole disk, but I couldn't find it. Perhaps it is somehow affected by the fact that 'is_whole_disk_image = true' is specified in the deployment process. I also have an x86_64 host as a compute node, it boots in legacy mode. I got the image for it from here - https://images.rdoproject.org/centos8/wallaby/rdo_trunk/current-tripleo/ I also get the same error on it. What could affect this? Full log of one of the nodes - https://pastebin.pl/view/5a6a7587 Node Show output of this node - https://pastebin.pl/view/ec5950b6

1 0

[URGENT] - Help needed with open stack services hung
by Karl Kloppenborg 12 Dec '23

12 Dec '23

Hi Openstack Community, Unfortunately we’ve had a weird situation arise where one of our regions has just stopped dead in it’s tracks, all management systems have been rebooted but there seems to be something stopping services from processing requests. Database, RabbitMQ Memcached have all been individually tested and I cannot find fault with them. Could any openstack consultants who are available for immediate emergency work please contact me off list. Thanks, Karl. Kind Regards, Karl Kloppenborg BCompSc, CNCF-[KCNA, CKA, CKAD], LFCE, CompTIA Linux+ XK0-004 Chief Technology Officer m: +61 437 239 565<mailto:+61%20437%20239%20565> | e: kkloppenborg(a)resetdata.com.au<mailto:kkloppenborg@resetdata.com.au> | w: https://resetdata.com [signature_2024530915][Image] This email transmission is intended only for the addressee / person responsible for delivery of the message to such person and may contain confidential or privileged information. Confidentiality and legal privilege are not waived or lost by reason of mistaken delivery to you, nor may you use, review, disclose, disseminate or copy any information contained in or attached to it. Whilst this email has been checked for viruses, the sender does not warrant that any attachments are free from viruses or other defects. You assume all liability for any loss, damage or other consequences which may arise from opening or using the attachments. If you received this e-mail in error please delete it and any attachments and kindly notify us by immediately sending an email to contact(a)resetdata.com.au<mailto:contact@resetdata.com.au>

1 0

Individual Board Director nominations: 3d left
by Kurt Garloff 12 Dec '23

12 Dec '23

Hi, the Open Infra communities have many strengths. One of them is a well-balanced board of directors that has representatives from all member classes, the Platinum members, the Gold members, and from individuals [1] that are part of the communities and elected -- each class contributing 1/3 of the seats.[4] I am worried about the individual class though: There are currently only 3 candidates that have enough nominations to be on the ballot [2]. 3 more needing a few more nominations. But I believe there should be a few more candidates on the ballot, otherwise it becomes an election without choice, something that is not good IMVHO. So please identify individuals that want to serve and nominate [3] them. This makes the elections meaningful and allows the board to do its work with also the individual members well represented. And there are only 3 more days to go for this! [1] https://openinfra.dev/election/2024-individual-director-election [2] https://openinfra.dev/election/2024-individual-director-election/candidates [3] https://openinfra.dev/a/community/members [4] https://openinfra.dev/about/board/ Thanks, -- Kurt Garloff <openstack(a)garloff.de>, Cologne CTO Sovereign Cloud Stack project @ OSB Alliance e.V.

2 1

[kolla] [2023.1] [masakari] Orphaned instances on node shutdown
by Rob Jefferson 11 Dec '23

11 Dec '23

Folks, I'm running into an issue with instances appearing to be running, even though they shouldn't be. The environment in this case is a virtualized OpenStack Kolla (2023.1) cluster, with three compute nodes and three control plane nodes. I am trying to see what happens when I create and terminate instances, in an attempt to figure out failover behavior. The three computes are in a single segment. I created four instances: +--------------------+--------+---------------------------+ | Name | Status | Host | +--------------------+--------+---------------------------+ | rj_testinstance_04 | ACTIVE | devtest-libvirt1-muon1005 | | rj_testinstance_03 | ACTIVE | devtest-libvirt1-muon1004 | | rj_testinstance_02 | ACTIVE | devtest-libvirt1-muon1006 | | rj_testinstance | ACTIVE | devtest-libvirt1-muon1005 | +--------------------+--------+---------------------------+ I decided to see what would happen if I went in and shutdown devtest-libvirt1-muon-1005 (which has two instances running). So I ssh'd into the machine and entered (as root): `echo c > /proc/sysrq-trigger` to force a crash, which duly occurred. I then made sure the host was seen as down: +---------------------------+-------+ | Hypervisor Hostname | State | +---------------------------+-------+ | devtest-libvirt1-muon1005 | down | | devtest-libvirt1-muon1004 | up | | devtest-libvirt1-muon1006 | up | +---------------------------+-------+ But even after this, the instances seem to still be up, including on the now crashed machine: +--------------------+--------+---------------------------+ | Name | Status | Host | +--------------------+--------+---------------------------+ | rj_testinstance_04 | ACTIVE | devtest-libvirt1-muon1005 | | rj_testinstance_03 | ACTIVE | devtest-libvirt1-muon1004 | | rj_testinstance_02 | ACTIVE | devtest-libvirt1-muon1006 | | rj_testinstance | ACTIVE | devtest-libvirt1-muon1005 | +--------------------+--------+---------------------------+ I would have expected the instances to have been regenerated on other compute hosts, or some sort of error. The nodes are on the same segment, so Masakari should have at least attempted to roll over the instances to other computes in the segment. But checking the instance details suggests that nothing has happened. The instances are just... lost. The Masakari host monitor log file reflects reality: 2023-12-11 19:21:49.953 8 INFO masakarimonitors.hostmonitor.host_handler.handle_host [-] 'devtest-libvirt1-muon1005' is 'offline' (current: 'offline'). Needless to say, I'm very confused here and am sure what's going on. Any advice would help. Thanks, Rob

1 1

[neutron][ovn] Outbound SNAT behavior
by James Denton 11 Dec '23

11 Dec '23

Hi everyone, While looking at an OVS->OVN migration, we came across a use-case that doesn’t provide 1:1 compatibility and I wanted to check with the community to see if there were any suggestions or ideas on how to proceed. In an OVS-based environment, we have users with a typical layout: <external provider network> | <neutron router> | | <tenantnet1> <tenantnet2> The users have implemented static routes on the Neutron router using the extraroutes API, which works great, and the default SNAT behavior implemented within the qrouter namespace handles any network traffic egressing the router, not just directly-connected tenant networks (see iptables entries below): -A neutron-l3-agent-snat -j neutron-l3-agent-float-snat -A neutron-l3-agent-snat -o qg-700f84b0-6b -j SNAT --to-source 168.47.160.223 --random-fully -A neutron-l3-agent-snat -m mark ! --mark 0x2/0xffff -m conntrack --ctstate DNAT -j SNAT --to-source 168.47.160.223 --random-fully -A neutron-postrouting-bottom -m comment --comment "Perform source NAT on outgoing traffic." -j neutron-l3-agent-snat OVN, on the other hand, appears to implement snat rules in the logical router based on directly-connected tenant networks, which means that traffic sourcing from other networks (not managed by Neutron but routable within a Neutron router, including to a VM or appliance) is not source NAT’d and egresses the router as its real IP. This behavior results in the traffic being effectively blackholed, as there’s no return route and no real desire to have exposed those IPs anyway. Here’s an example of the nat list which includes the directly connected tenant network: root@lab-controller01:~# ovn-nbctl lr-nat-list neutron-edfd7115-5a81-46b7-9692-fe227d395b0b TYPE GATEWAY_PORT EXTERNAL_IP EXTERNAL_PORT LOGICAL_IP EXTERNAL_MAC LOGICAL_PORT snat 192.168.3.33 10.254.254.0/24 The extraroute in the router is: destination: 172.16.0.0/24, nexthop: 10.254.254.6 I can add an entry manually, which allows the snat for the secondary network to work as expected, but results in some other issues where the snat(s) don’t get added back on a disable/enable of the router: root@lab-controller01:~# ovn-nbctl lr-nat-add neutron-edfd7115-5a81-46b7-9692-fe227d395b0b snat 192.168.3.33 172.16.0.0/24 root@lab-controller01:~# ovn-nbctl lr-nat-list neutron-edfd7115-5a81-46b7-9692-fe227d395b0b TYPE GATEWAY_PORT EXTERNAL_IP EXTERNAL_PORT LOGICAL_IP EXTERNAL_MAC LOGICAL_PORT snat 192.168.3.33 10.254.254.0/29 snat 192.168.3.33 172.16.0.0/24 I understand this is probably not a supported usecase, but one that worked nonetheless, so any suggestions on how to mirror this functionality in a supported way would be greatly appreciated. Thanks, James

1 0

[tc] Technical Committee Meeting Tomorrow, Dec 12
by Jay Faulkner 11 Dec '23

11 Dec '23

Hi all, The weekly meeting of the Technical Committee will be held Tuesday, Dec 12 at 1800 UTC in #openstack-tc on OFTC. See below for the proposed agenda: - Roll call - Follow up on tracked action items: https://meetings.opendev.org/meetings/tc/2023/tc.2023-12-05-18.00.html - Gmann - Create TC resolution expressing approval for OIF bylaws change - Create TC charter change in response to OIF bylaws changes - JayF - Communicate deadline for approval of above OIF bylaws related governance changes - Gate health check - Leaderless projects (gmann) - Implementation of Unmaintained branch statuses - 2024.1 TC Tracker - https://etherpad.opendev.org/p/tc-2024.1-tracker - Open Discussion and Reviews Thanks all, Jay Faulkner TC Chair

1 0

[kolla-ansible] rocky spice containers missing
by Michael Knox 11 Dec '23

11 Dec '23

Hi all. I was looking to use SPICE instead of VNC, but noticed that there are no rocky containers for it on quay.io. I found a ticket for CentOS 8 source, is this still the same reason? https://bugs.launchpad.net/kolla-ansible/+bug/1955549 Cheers Michael

3 3

[neutron] Bug Deputy Report December 4 - December 10
by Lajos Katona 11 Dec '23

11 Dec '23

Hi Neutrinos, I was the Neutron bug deputy last week, and I send the report of last week's bugs. - High prio bugs - Unassigned: https://bugs.launchpad.net/neutron/+bug/2045549 (OVS jobs randomly fails as Guest VMs not(or delayed) configured with DHCP) - https://bugs.launchpad.net/neutron/+bug/2045725 (scenario jobs randomly fails downloading ubuntu image) - related devstack patch: https://review.opendev.org/c/openstack/devstack/+/901699 - Medium prio bugs: - https://bugs.launchpad.net/neutron/+bug/2045811 (neutron-ovn-db-sync-util can fail with KeyError) - https://review.opendev.org/c/openstack/neutron/+/902841 - https://bugs.launchpad.net/neutron/+bug/2045889 ([OVN] ML2/OVN mech driver does not set the OVS bridge name in the port VIF details) - related mail thread: https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack… - https://bugs.launchpad.net/neutron/+bug/2045950 (Security group performance issue for iptables driver due to "stateless feature") - performance issue with stateless security-groups. Best wishes Lajos Katona (lajoskatona)

1 0

[all] Ussuri and Train EOL
by Elõd Illés 11 Dec '23

11 Dec '23

Hi, With the new TC resolution '2023-07-24 Unmaintained status replaces Extended Maintenance' [1] the last 3 EM branches (Xena, Wallaby, Victoria) will transition to Unmaintained, but older branches (Ussuri, Train) should move to End of Life unless someone steps up as maintainer of this branch for this project. (In that case the branch of that project will transition to unmaintained/{ussuri,train}) To signal this, please add a -1 for the given generated EOL patch [2][3] if you or the team want to keep this branch open for further maintenance. If there is no response from anyone within a reasonable time frame, then PTLs/release liaisons will be asked to confirm the EOL transition with their +1. Thanks, Előd Illés irc: elodilles @ #openstack-stable / #openstack-release [1] https://governance.openstack.org/tc/resolutions/20230724-unmaintained-branc… [2] https://review.opendev.org/q/topic:train-eol [3] https://review.opendev.org/q/topic:ussuri-eol

1 0